Skip to content

Analysis: QA error on Release environment

Context

Today, while performing the preparation for the next patch release, it was noticed that QA execution on the release environment 18.1 was failing

Screenshot_2025-06-23_at_11.33.41_a.m.

QA job example: https://gitlab.com/gitlab-org/security/gitlab/-/jobs/10438927219

Failure/Error: raise Capybara::ElementNotFound, "Unable to find #{query.applied_description}" if result.empty?
Capybara::ElementNotFound:
  Unable to find css "[data-testid=\"access-token-name-field\"]"

The DevEx DRI pinpointed the failure:

It looks like it might be https://gitlab.com/gitlab-org/gitlab/-/blob/master/app/assets/javascripts/vue_shared/access_tokens/components/access_token_form.vue that is being used now instead which doesn't have the data-testid 🤔

I saw that there's a id="token-create-form" in the artifact's DOM and looked that up in the codebase and this was the only file with that id 11:24 Looks like it might be the culprit gitlab-org/gitlab#387072 (closed) Form was recreated in Vue and the data-testid was likely missed

Problem

The associated MR was merged on 2025-067-10 gitlab-org/gitlab!193648 (merged), and the QA failure was discovered 2025-06-23, more than one week later. In the middle:

However, none of the above detected the failure. The failure was spotted on the first deployment of the release environment.

This issue is to understand the quality gap. Some questions that might help us here:

  • Why this problem wasn't caught at the merge request level?
  • Why was this issue caught so late, by the release environments?
  • Why was this issue not detected by the QA executed on the pre or release environments?

Follow-up actions

Edited by Mayra Cabrera